{"id":"ef025619ba8f5690cd12dca80ce4fd9aecf786131b5068ad8f9e119ac5750650","title":"METR explains how task substitution changes measures of AI uplift","argument":"METR’s article was published May 8, 2026. It argues that estimates of AI’s productivity impact depend on which tasks are counted. METR distinguishes uplift on old tasks, uplift on new tasks, and uplift in value after people shift between tasks. Under simplifying assumptions, uplift on old tasks is no greater than uplift in value, which is no greater than uplift on new tasks. The distinction matters because people may choose tasks that AI makes much faster, even when those tasks add less value than their apparent time savings suggest. METR calls a task a “Cadillac Task” when someone does it because AI made it cheap and its uplift in value is much lower. Applying this framework, METR says an estimate based on old tasks might be a lower bound, but argues that the Claude-based estimate it discusses is likely an overestimate: speedups on selected tasks may not represent all tasks in the same task group. METR’s conclusion is that high uplift on new tasks need not mean equally high uplift in value. The inequalities depend on assumptions about people’s beliefs and choices; practical factors can make them fail. The article does not measure how much substitution within task groups occurs or how much it changes estimates of value uplift.","topicIds":["work-and-business--jobs-and-occupations","job-displacement-and-creation"],"domain":"other","claimType":"attributed reading","limits":"Text as retrieved on October 23, 2026; originally published May 8, 2026. METR says the inequalities require simplifying assumptions and that practical difficulties make uplift challenging to estimate. The article gives no empirical measure of how much substitution within task groups occurs in practice.","speaker":{"name":"METR","roleAtSource":"Publisher"},"source":{"title":"Task Substitution and Uplift","publisher":"METR","format":"Article","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","publishedAt":"2026-05-08T07:00:00.000Z","dateBasis":"Date supplied by the publisher’s feed."},"evidence":[{"text":"METR distinguishes three measures of AI productivity uplift based on whether the calculation uses old tasks, new tasks, or value after task reshuffling.","excerpt":"We have found it important to distinguish between three measures of AI’s uplift:","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"Under simplifying assumptions, METR ranks uplift on old tasks at or below uplift in value, and uplift in value at or below uplift on new tasks.","excerpt":"Under some simplifying assumptions, the following inequalities hold:","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR says the bounds rely on decision-makers allocating time to maximize an objective while holding accurate beliefs about task-level uplift.","excerpt":"These inequalities hold for decision-makers who allocate their time across tasks to maximize some objective, given accurate beliefs about task-level uplift.","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR argues that the Claude productivity estimate it discusses may overstate uplift in value because observed speedups on specific tasks may not represent all tasks in their task groups.","excerpt":"If these time-savings were representative of all tasks in that ONET task group (both pre-AI and post-AI) then the conclusion would be valid.","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR uses “Cadillac Task” for a task someone does because AI made it cheap when its uplift in value is much lower.","excerpt":"In the case where the uplift in value is much lower and the person is only doing the task because AI has made it cheap, we have been calling it a “Cadillac Task”.","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR says practical factors can make the inequalities fail in practice.","excerpt":"Some of these considerations could cause the inequalities to reverse in practice","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"}],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","checkedAt":null,"items":[]},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How much substitution within task groups occurs in practice, and how much does it change estimates of uplift in value?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"METR explains how task substitution changes measures of AI uplift","summary":"METR’s article was published May 8, 2026. It argues that estimates of AI’s productivity impact depend on which tasks are counted. METR distinguishes uplift on old tasks, uplift on new tasks, and uplift in value after people shift between tasks. Under simplifying assumptions, uplift on old tasks is no greater than uplift in value, which is no greater than uplift on new tasks. The distinction matters because people may choose tasks that AI makes much faster, even when those tasks add less value than their apparent time savings suggest. METR calls a task a “Cadillac Task” when someone does it because AI made it cheap and its uplift in value is much lower. Applying this framework, METR says an estimate based on old tasks might be a lower bound, but argues that the Claude-based estimate it discusses is likely an overestimate: speedups on selected tasks may not represent all tasks in the same task group. METR’s conclusion is that high uplift on new tasks need not mean equally high uplift in value. The inequalities depend on assumptions about people’s beliefs and choices; practical factors can make them fail. The article does not measure how much substitution within task groups occurs or how much it changes estimates of value uplift.","question":"How much substitution within task groups occurs in practice, and how much does it change estimates of uplift in value?","claims":[{"text":"METR distinguishes three measures of AI productivity uplift based on whether the calculation uses old tasks, new tasks, or value after task reshuffling.","excerpt":"We have found it important to distinguish between three measures of AI’s uplift:","support":"source-stated"},{"text":"Under simplifying assumptions, METR ranks uplift on old tasks at or below uplift in value, and uplift in value at or below uplift on new tasks.","excerpt":"Under some simplifying assumptions, the following inequalities hold:","support":"source-stated"},{"text":"METR says the bounds rely on decision-makers allocating time to maximize an objective while holding accurate beliefs about task-level uplift.","excerpt":"These inequalities hold for decision-makers who allocate their time across tasks to maximize some objective, given accurate beliefs about task-level uplift.","support":"source-stated"},{"text":"METR argues that the Claude productivity estimate it discusses may overstate uplift in value because observed speedups on specific tasks may not represent all tasks in their task groups.","excerpt":"If these time-savings were representative of all tasks in that ONET task group (both pre-AI and post-AI) then the conclusion would be valid.","support":"source-stated"},{"text":"METR uses “Cadillac Task” for a task someone does because AI made it cheap when its uplift in value is much lower.","excerpt":"In the case where the uplift in value is much lower and the person is only doing the task because AI has made it cheap, we have been calling it a “Cadillac Task”.","support":"source-stated"},{"text":"METR says practical factors can make the inequalities fail in practice.","excerpt":"Some of these considerations could cause the inequalities to reverse in practice","support":"source-stated"}],"limitations":["Text as retrieved on October 23, 2026; originally published May 8, 2026.","METR says the inequalities require simplifying assumptions and that practical difficulties make uplift challenging to estimate.","The article gives no empirical measure of how much substitution within task groups occurs in practice."],"mediaId":null,"correctionReason":null,"id":"ef025619ba8f5690cd12dca80ce4fd9aecf786131b5068ad8f9e119ac5750650","publicationId":"f9256cbb-4cd2-b95b-5f3d-b8cb471db8d3","version":1,"sourceUrl":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","sourceTitle":"Task Substitution and Uplift","publisher":"METR","publishedAt":"2026-05-08T07:00:00.000Z","originalPublished":{"value":"2026-05-08T07:00:00.000Z","precision":"instant","basis":"backfill:feed_published"},"firstObservedAt":"2026-10-18T02:55:23.126Z","createdAt":"2026-11-07T03:33:23.866Z","historical":true,"entityIds":["metr"],"topicIds":["work-and-business--jobs-and-occupations","job-displacement-and-creation"],"media":[],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","checkedAt":null,"items":[]},"editorialDecision":{"outcome":"accept","mediaId":null,"inputHash":"0fc62f04270ed77cc9fa083ddf5f9951a2433367d4009a7377d39565bf08fca9","rubricVersion":"unrecorded","provider":"vercel","model":"typesafe-ai/jev","reasonCodes":["bridge_admitted_by_model"],"confidence":0.87,"decidedAt":"2026-10-23T22:28:01.951Z"},"provenance":{"textHash":"b960b6a416c9610b27e1809d114024456d110684aaea3d3f66009cae55ac93a2","observationId":"c1d0a6d300eb6d24fba11471f605be306994044439b48c5de7a23fa85fe19799","retrievedAt":"2026-10-23T22:41:38.132Z","kind":"publisher-page","preparer":{"model":"openai/gpt-6-luna","runId":"4c3f3bd8-13f9-5c7f-7122-91c6622ef54b"},"reviewer":{"model":"anthropic/claude-opus-5.5","runId":"17292420-c335-1627-c23f-7cd586826bd1"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-07T03:28:29.149Z","model":"anthropic/claude-opus-5.5"}},"canonicalUrl":"https://escapemeridian.com/readings/metr-explains-how-task-substitution-changes-measures-of-ai-ef025619","formats":{"html":"/readings/metr-explains-how-task-substitution-changes-measures-of-ai-ef025619","markdown":"/readings/metr-explains-how-task-substitution-changes-measures-of-ai-ef025619.md","json":"/readings/metr-explains-how-task-substitution-changes-measures-of-ai-ef025619.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[],"entities":[{"id":"metr","name":"METR","kind":"organization","context":"METR studies what AI systems can do and how reliably they complete tasks, including work with human developers.","referenceUrls":["https://metr.org/"],"path":"/entities/metr","json":"/entities/metr.json","markdown":"/entities/metr.md"},{"id":"time","name":"TIME","kind":"organization","context":"TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.","referenceUrls":["https://time.com/"],"path":"/entities/time","json":"/entities/time.json","markdown":"/entities/time.md"},{"id":"claude","name":"Claude","kind":"model","context":"Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.","referenceUrls":["https://claude.com/product/overview","https://www.anthropic.com/news","https://www.anthropic.com/research/team/economics"],"path":"/entities/claude","json":"/entities/claude.json","markdown":"/entities/claude.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"ef025619ba8f5690cd12dca80ce4fd9aecf786131b5068ad8f9e119ac5750650","revision":"f9256cbb-4cd2-b95b-5f3d-b8cb471db8d3","revisionKind":"publication-id","title":"METR explains how task substitution changes measures of AI uplift","url":"https://escapemeridian.com/readings/metr-explains-how-task-substitution-changes-measures-of-ai-ef025619"},"source":{"type":"source","id":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","revision":"b960b6a416c9610b27e1809d114024456d110684aaea3d3f66009cae55ac93a2","revisionKind":"sha256-source-text","title":"Task Substitution and Uplift","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"ef025619ba8f5690cd12dca80ce4fd9aecf786131b5068ad8f9e119ac5750650","title":"METR explains how task substitution changes measures of AI uplift","argument":"METR’s article was published May 8, 2026. It argues that estimates of AI’s productivity impact depend on which tasks are counted. METR distinguishes uplift on old tasks, uplift on new tasks, and uplift in value after people shift between tasks. Under simplifying assumptions, uplift on old tasks is no greater than uplift in value, which is no greater than uplift on new tasks. The distinction matters because people may choose tasks that AI makes much faster, even when those tasks add less value than their apparent time savings suggest. METR calls a task a “Cadillac Task” when someone does it because AI made it cheap and its uplift in value is much lower. Applying this framework, METR says an estimate based on old tasks might be a lower bound, but argues that the Claude-based estimate it discusses is likely an overestimate: speedups on selected tasks may not represent all tasks in the same task group. METR’s conclusion is that high uplift on new tasks need not mean equally high uplift in value. The inequalities depend on assumptions about people’s beliefs and choices; practical factors can make them fail. The article does not measure how much substitution within task groups occurs or how much it changes estimates of value uplift.","topicIds":["work-and-business--jobs-and-occupations","job-displacement-and-creation"],"domain":"other","claimType":"attributed reading","limits":"Text as retrieved on October 23, 2026; originally published May 8, 2026. METR says the inequalities require simplifying assumptions and that practical difficulties make uplift challenging to estimate. The article gives no empirical measure of how much substitution within task groups occurs in practice.","speaker":{"name":"METR","roleAtSource":"Publisher"},"source":{"title":"Task Substitution and Uplift","publisher":"METR","format":"Article","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","publishedAt":"2026-05-08T07:00:00.000Z","dateBasis":"Date supplied by the publisher’s feed."},"evidence":[{"text":"METR distinguishes three measures of AI productivity uplift based on whether the calculation uses old tasks, new tasks, or value after task reshuffling.","excerpt":"We have found it important to distinguish between three measures of AI’s uplift:","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"Under simplifying assumptions, METR ranks uplift on old tasks at or below uplift in value, and uplift in value at or below uplift on new tasks.","excerpt":"Under some simplifying assumptions, the following inequalities hold:","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR says the bounds rely on decision-makers allocating time to maximize an objective while holding accurate beliefs about task-level uplift.","excerpt":"These inequalities hold for decision-makers who allocate their time across tasks to maximize some objective, given accurate beliefs about task-level uplift.","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR argues that the Claude productivity estimate it discusses may overstate uplift in value because observed speedups on specific tasks may not represent all tasks in their task groups.","excerpt":"If these time-savings were representative of all tasks in that ONET task group (both pre-AI and post-AI) then the conclusion would be valid.","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR uses “Cadillac Task” for a task someone does because AI made it cheap when its uplift in value is much lower.","excerpt":"In the case where the uplift in value is much lower and the person is only doing the task because AI has made it cheap, we have been calling it a “Cadillac Task”.","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"},{"text":"METR says practical factors can make the inequalities fail in practice.","excerpt":"Some of these considerations could cause the inequalities to reverse in practice","url":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","relationship":"Source-stated"}],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","checkedAt":null,"items":[]},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How much substitution within task groups occurs in practice, and how much does it change estimates of uplift in value?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"METR explains how task substitution changes measures of AI uplift","summary":"METR’s article was published May 8, 2026. It argues that estimates of AI’s productivity impact depend on which tasks are counted. METR distinguishes uplift on old tasks, uplift on new tasks, and uplift in value after people shift between tasks. Under simplifying assumptions, uplift on old tasks is no greater than uplift in value, which is no greater than uplift on new tasks. The distinction matters because people may choose tasks that AI makes much faster, even when those tasks add less value than their apparent time savings suggest. METR calls a task a “Cadillac Task” when someone does it because AI made it cheap and its uplift in value is much lower. Applying this framework, METR says an estimate based on old tasks might be a lower bound, but argues that the Claude-based estimate it discusses is likely an overestimate: speedups on selected tasks may not represent all tasks in the same task group. METR’s conclusion is that high uplift on new tasks need not mean equally high uplift in value. The inequalities depend on assumptions about people’s beliefs and choices; practical factors can make them fail. The article does not measure how much substitution within task groups occurs or how much it changes estimates of value uplift.","question":"How much substitution within task groups occurs in practice, and how much does it change estimates of uplift in value?","claims":[{"text":"METR distinguishes three measures of AI productivity uplift based on whether the calculation uses old tasks, new tasks, or value after task reshuffling.","excerpt":"We have found it important to distinguish between three measures of AI’s uplift:","support":"source-stated"},{"text":"Under simplifying assumptions, METR ranks uplift on old tasks at or below uplift in value, and uplift in value at or below uplift on new tasks.","excerpt":"Under some simplifying assumptions, the following inequalities hold:","support":"source-stated"},{"text":"METR says the bounds rely on decision-makers allocating time to maximize an objective while holding accurate beliefs about task-level uplift.","excerpt":"These inequalities hold for decision-makers who allocate their time across tasks to maximize some objective, given accurate beliefs about task-level uplift.","support":"source-stated"},{"text":"METR argues that the Claude productivity estimate it discusses may overstate uplift in value because observed speedups on specific tasks may not represent all tasks in their task groups.","excerpt":"If these time-savings were representative of all tasks in that ONET task group (both pre-AI and post-AI) then the conclusion would be valid.","support":"source-stated"},{"text":"METR uses “Cadillac Task” for a task someone does because AI made it cheap when its uplift in value is much lower.","excerpt":"In the case where the uplift in value is much lower and the person is only doing the task because AI has made it cheap, we have been calling it a “Cadillac Task”.","support":"source-stated"},{"text":"METR says practical factors can make the inequalities fail in practice.","excerpt":"Some of these considerations could cause the inequalities to reverse in practice","support":"source-stated"}],"limitations":["Text as retrieved on October 23, 2026; originally published May 8, 2026.","METR says the inequalities require simplifying assumptions and that practical difficulties make uplift challenging to estimate.","The article gives no empirical measure of how much substitution within task groups occurs in practice."],"mediaId":null,"correctionReason":null,"id":"ef025619ba8f5690cd12dca80ce4fd9aecf786131b5068ad8f9e119ac5750650","publicationId":"f9256cbb-4cd2-b95b-5f3d-b8cb471db8d3","version":1,"sourceUrl":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","sourceTitle":"Task Substitution and Uplift","publisher":"METR","publishedAt":"2026-05-08T07:00:00.000Z","originalPublished":{"value":"2026-05-08T07:00:00.000Z","precision":"instant","basis":"backfill:feed_published"},"firstObservedAt":"2026-10-18T02:55:23.126Z","createdAt":"2026-11-07T03:33:23.866Z","historical":true,"entityIds":["metr"],"topicIds":["work-and-business--jobs-and-occupations","job-displacement-and-creation"],"media":[],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2026-05-08-task-substitution-and-uplift/","checkedAt":null,"items":[]},"editorialDecision":{"outcome":"accept","mediaId":null,"inputHash":"0fc62f04270ed77cc9fa083ddf5f9951a2433367d4009a7377d39565bf08fca9","rubricVersion":"unrecorded","provider":"vercel","model":"typesafe-ai/jev","reasonCodes":["bridge_admitted_by_model"],"confidence":0.87,"decidedAt":"2026-10-23T22:28:01.951Z"},"provenance":{"textHash":"b960b6a416c9610b27e1809d114024456d110684aaea3d3f66009cae55ac93a2","observationId":"c1d0a6d300eb6d24fba11471f605be306994044439b48c5de7a23fa85fe19799","retrievedAt":"2026-10-23T22:41:38.132Z","kind":"publisher-page","preparer":{"model":"openai/gpt-6-luna","runId":"4c3f3bd8-13f9-5c7f-7122-91c6622ef54b"},"reviewer":{"model":"anthropic/claude-opus-5.5","runId":"17292420-c335-1627-c23f-7cd586826bd1"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-07T03:28:29.149Z","model":"anthropic/claude-opus-5.5"}}}],"audits":[{"publicationId":"f9256cbb-4cd2-b95b-5f3d-b8cb471db8d3","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The reading attributes the framework, inequalities, critique of the Claude estimate and the term Cadillac Task to METR, the publisher of this single blog post. The arguments stay framed as METR's claims.","excerpt":"We have found it important to distinguish between three measures of AI’s uplift:"},{"dimension":"summary","status":"uncertain","explanation":"The summary is accurate but leaves central terms hard for a non-specialist to follow, and no definitions are supplied. It never says that old tasks means a typical pre-AI workday task mix and new tasks means the mix chosen once AI is available. It does not explain that task group refers to O*NET task categories. It also leaves the Claude-based estimate unidentified; the source attributes it to Tamkin and McCrory (2025), at roughly 17%. Smallest repair: one plain-language clause for each of these terms and a short identification of that estimate.","excerpt":"“New tasks” means the set of tasks you’d do in a typical day after AI is available."},{"dimension":"date","status":"pass","explanation":"The reading gives May 8, 2026, which matches the UTC day of evidence.originalPublished and the URL path. The evidence kind is publisher-page, so no archive disclosure is needed. The reading presents no later events, uses no hindsight framing and does not rely on the source's present-tense remark about coding agents.","excerpt":null},{"dimension":"support","status":"pass","explanation":"Each claim matches the source. The inequalities are tied to simplifying assumptions and to maximizing behavior with accurate beliefs. The source says the old-task estimate is an expected lower bound but likely an overestimate because of within-group substitution, and the reading reports both points. Cadillac Task is defined as in the source, and the reading notes that practical factors could reverse the inequalities. The stated limitation about an unmeasured extent of substitution fairly reflects the article.","excerpt":"However it seems likely there is substantial substitution within task groups, towards those which have higher time-savings."},{"dimension":"media","status":"pass","explanation":"No media was selected and no verified media candidates were supplied, so no media identity or relevance issue arises.","excerpt":null}],"createdAt":"2026-11-07T03:28:29.149Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}